Multislice Computed Tomography in the Diagnosis of Superior Canal Dehiscence: How Much Error, and How to Minimize It? *†Tanya S. Tavassolie, †‡§Richard T. Penninger, †kM. Geraldine Zun ˜iga, †Lloyd B. Minor, and †John Patrick Carey *Franklin and Marshall College, Lancaster, Pennsylvania; ÞThe Johns Hopkins University School of Medicine, Baltimore, Maryland, U.S.A.; þUpper Austria University of Applied Sciences, Linz, Austria; §Ghent University, Ghent, Belgium; and kEscuela de Medicina Ignacio A. Santos del Tecnolo ´gico de Monterrey, Mexico Hypothesis: Multi-slice computed tomography (MSCT) over- estimates the size of superior semicircular canal dehiscences (SSCDs) and also can misinterpret thin bone over the superior semicircular canal as dehiscent. A threshold of the radiodensity of the bone over the superior semicircular canal may exist that could optimize prediction of an actual SSCD. Background: The gold standard for diagnosis of SSCD is MSCT, but there is a higher prevalence of SSCD based on MSCT compared with histologic studies. Overestimation of SSCD can lead to inappropriate diagnosis and treatment. Methods: We correlated radiographic and surgical findings in SSCD to determine if MSCT overestimated the size of SSCD and if a threshold radiodensity could be defined, below which actual dehiscence could best be predicted. Participants were 34 humans with SSCD confirmed at surgery. MSCT scans were acquired axially with 0.5-mm collimation and a small field of view (24 cm). Dehiscence sizes measured from radial recon- structions were compared with measurements made during surgery. Results: There were significant differences between radiographic and actual length and width, indicating that MSCT tends to overestimate the size of SSCD. Receiver operating characteristic analysis found a threshold in Hounsfield units that optimized the prediction of dehiscence. Conclusion: Computed tomographic imaging alone can be mis- leading for diagnosis of SSCD. It can overestimate the size of the dehiscence, and it can falsely detect dehiscences. Clinical symptoms and other signs must be clearly indicative before surgery, and MSCT cannot be used exclusively for the diagno- sis of SSCD. Key Words: Computed Labyrinth Receiver op- erating characteristicVSensitivityVSpecificityVSuperior canal dehiscenceVVertigoVVestibular. Otol Neurotol 33:215Y222, 2012. Superior semicircular canal dehiscence (SSCD) is a bony dehiscence of the superior semicircular canal into the intracranial space that, when symptomatic, presents with auditory and/or vestibular symptoms, collectively known as SSCD syndrome (SSCDS) (1). Auditory symp- toms include autophony, aural fullness, and conductive hyperacusis (2Y4). Vestibular symptoms include dis- equilibrium or vertigo induced by changes in intracranial or middle-ear pressure or by sound. Diagnosis of SSCD is confirmed by a combination of imaging, audiometric testing, vestibular-evoked myogenic potential (VEMP) testing, and physical examination (1,2,4,5). MSCT is presently considered the ideal imaging mo- dality for SSCD. However, the resolution of MSCT for the thin layer of bone that may overlay the superior canal (SC) has not been determined. This resolution is affected by several parameters. Collimation of the x-ray beam is among the most important of these. Although conven- tional temporal bone MSCT scans with 1.0-mm collima- tion viewed in the axial and coronal planes have excellent sensitivity for detecting SSCD, the specificity and positive predictive value of the imaging study can be improved with 0.5-mm collimation and reconstructions in the plane of the SC (4). Field of view also affects resolution, and it is best to map the smallest volume of tissue to the fixed image matrix size to minimize partial volume averaging and improve resolution. Filtering of the raw MSCT data to produce meaningful images is typically done with an edge detection filter, but the noise reduction algorithms may effectively remove thin bone from the final image, Address correspondence and reprint requests to John Patrick Carey, M.D., Department of Otolaryngology-Head and Neck Surgery, Johns Hopkins Outpatient Center, 601 North Caroline Street, 6th Floor, Baltimore, MD 21287. E-mail: jcarey@jhmi.edu The authors declare no conflicts of interest. Otology & Neurotology 33:215Y222 Ó 2012, Otology & Neurotology, Inc. 215 Copyright © 2012 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.